famous Quantity Theory of Money. So long as _k_ remains unchanged, _n_ and _p_ rise and fall together; that is to say, the greater or the fewer the number of currency notes, the higher or the lower is the price level in the same proportion. So far we have assumed that the whole of the public requirement for purchasing power is satisfied by cash, and on the other hand that this requirement is the only source of demand for cash; neglecting the fact that the public, including the business world, employ for the same purpose bank deposits and overdraft facilities, whilst the banks must for the same reason maintain a reserve of cash. The theory is easily extended, however, to cover this case. Let us assume that the public, including the business world, find it convenient to keep the equivalent of _k_ consumption units in cash and of a further _k´_ available at their banks against cheques, and that the banks keep in cash a proportion _r_ of their potential liabilities (_k´_) to the public. Our equation then becomes _n = p(k + rk´)_. So long as _k_, _k´_, and _r_ remain unchanged, we have the same result as before, namely, that _n_ and _p_ rise and fall together. The proportion between _k_ and _k´_ depends on the banking arrangements of the public; the absolute value of these on their habits generally; and the value of _r_ on the reserve practices of the banks. Thus, so long as these are unaltered, we still have a direct relation between the _quantity_ of cash (_n_) and the level of prices (_p_).[21] [21] My exposition follows the general lines of Prof. Pigou (_Quarterly Journal of Economics_, Nov. 1917) and of Dr. Marshall (_Money, Credit, and Commerce_, I. iv.), rather than the perhaps more familiar analysis of Prof. Irving Fisher. Instead of starting with the amount of cash held by the public, Prof. Fisher begins with the volume of business transacted by means of money and the frequency with which each unit of money changes hands. It comes to the same thing in the end and it is easy to pass from the above formula to Prof. Fisher’s; but the above method of approach seems less artificial than Prof. Fisher’s and nearer to the observed facts. We have seen that the amount of _k_ and _k´_ depends partly on the wealth of the community, partly on its habits. Its habits are fixed by its estimation of the extra convenience of having more cash in hand as compared with the advantages to be got from spending the cash or investing it. The point of equilibrium is reached where the estimated advantages of keeping more cash in hand compared with those of spending or investing it about balance. The matter cannot be summed up better than in the words of Dr. Marshall: